4 ms·
It is an LLM fine-tuned using a new type of dataset and RL reward. It's good at reasoning, but I would not recommend to replace Llama for general tasks.
by peakji 2y ago
It is an LLM fine-tuned using a new type of dataset and RL reward. It's good at reasoning, but I would not recommend to replace Llama for general tasks.